Obsidian
About
Interact with your Obsidian vault from your IDE or Claude Desktop.
Details
- Author
- vasallo94
- Categories
- Productivity, Other, AI
Jump to
Setup
Install Obsidian in your MCP client (Claude Desktop, Cursor, Windsurf, and others).
Repository: https://github.com/vasallo94/obsidian-mcp-server
Follow the installation instructions in the repository README, then restart your MCP client.
AnMCP (Model Context Protocol)server that lets AI agents work inside an Obsidian vault: read notes, search context, inspect links, follow vault-specific rules, and optionally create or edit notes safely.
It is designed for clients and harnesses such asCodex,Claude Code,Hermes, andClaude Desktop. The core stays reusable; each vault can layer its own profiles, rules, skills, and optional tool sets on top.
Tools are generic. Behavior comes from the vault.
flowchart LR Clients["Codex, Claude Code, Hermes, Claude Desktop"] --> MCP["Obsidian MCP Server"] MCP --> Core["Core tools: read, search, inspect, route"] MCP --> Optional["Optional tool sets: write, graph, canvas, ObsidianRAG"] Core --> Vault["Obsidian vault"] Optional --> Vault Vault --> Profile[".agents/vault.yaml, rules, skills, standards"] Profile --> MCP
The core tool set is always available and stays vault-agnostic:
- Vault diagnostics, task routing, and MCP client root inspection.
- Note listing, reading, metadata inspection, and search.
- Vault context resources for profiles, skills, standards, and local docs.
- Core prompts for structured notes, template usage, and context exploration.
Optional packs are enabled explicitly from.agents/vault.yamlorOBSIDIAN_MCP_TOOL_SETS:
- notes_write: Create, patch, move, and delete notes.
- vault_analysis: Vault statistics, tags, links, backlinks, and graph tools.
- agents_admin: Skill creation, validation, and cache management.
- youtube: Transcript extraction.
- obsidianrag: Semantic search through the external ObsidianRAG service.
- canvas/kanvas: Canvas and workflow helpers.
- Profile packs: Personal workflows only when a vault profile opts in.
- Public core, personal profiles: The repository remains reusable; local workflows live in vault configuration and resources.
- English technical surface: Tool names, prompt names, docs, and code identifiers are English.
- Safe by default: Write tools are opt-in, protected paths are blocked, and large reads are capped.
- External RAG by integration: Advanced semantic search delegates to ObsidianRAG instead of duplicating a RAG stack inside the MCP server.
- uv
- An Obsidian vault path you are comfortable exposing to an MCP client
Until the package is published to PyPI, install directly from GitHub withuvx:
uvx --from git+https://github.com/Vasallo94/obsidian-mcp-server.git obsidian-mcp-server
For Codex, add this to~/.codex/config.toml:
[mcp_servers.obsidian] command = "uvx" args = [ "--from", "git+https://github.com/Vasallo94/obsidian-mcp-server.git", "obsidian-mcp-server", ] startup_timeout_sec = 30 tool_timeout_sec = 120 [mcp_servers.obsidian.env] OBSIDIAN_VAULT_PATH = "/absolute/path/to/your/vault"
For Claude Code, Hermes, Claude Desktop, and MCPB setup, seeInstallation.
git clone https://github.com/Vasallo94/obsidian-mcp-server.git cd obsidian-mcp-server make install cp .env.example .env # Set OBSIDIAN_VAULT_PATH to the absolute path to your Obsidian vault uv run obsidian-mcp-server
Once the package is published to PyPI, client configs can use:
Enable optional tools from the client environment:
{ "env": { "OBSIDIAN_VAULT_PATH": "/Absolute/Path/To/Your/Vault", "OBSIDIAN_MCP_TOOL_SETS": "notes_write,vault_analysis,obsidianrag" } }
profile: name: "my_profile" prompt_sets: - "mermaid" tool_sets: - "notes_write" - "vault_analysis" standards: media: "Standards/Media.md" local_docs: index: "README.md"
For semantic vault search, enable theobsidianragtool set and declare the integration:
profile: tool_sets: - "obsidianrag" integrations: obsidianrag: project_path: "/path/to/ObsidianRAG" api_url: "http://127.0.0.1:8000" env: OBSIDIANRAG_LLM_MODEL: "gemma3" OBSIDIANRAG_OLLAMA_EMBEDDING_MODEL: "embeddinggemma"
Then readobsidian://integrations/obsidianrag/setupor callrag.setup_status. Agents should show setup commands before installing dependencies, starting services, pulling models, or rebuilding the index.
To dive deeper into how the server works and how to customize it, check our detailed guides located in thedocs/folder:
- Documentation Home: Wiki-style map of the project docs.
- Installation: Setup for Codex, Claude Code, Hermes, Claude Desktop, and MCPB.
- Architecture: Runtime architecture, tool sets, resources, prompts, and security model.
- Tool Reference: Complete list of public MCP tools.
- Server Configuration: Environment variables, vault profiles, tool sets, and integrations.
- Agent Setup: How to organize your vault (.agents/) with skills and contextual rules.
- Semantic Search: ObsidianRAG integration and legacy RAG migration notes.
- Agent Feedback: How agents can report MCP friction with AFP out-of-band.
- Future Roadmap: Planned improvements and next steps for the server.
For contribution, release, and security process, seeCONTRIBUTING.md,SECURITY.md, andRelease Checklist.
This project is licensed under the MIT License.
Persistent memory for any AI assistant. Zero token cost until recall. Stores memories in local SQLite, ranks by 6-factor scoring, returns results 79% smaller than JSON. Works with Claude, ChatGPT, Grok, Cursor, Windsurf, and any MCP client.
After Effects MCP is a full-featured automation bridge that connects AI clients (like VS Code, Claude Desktop, and Claude Code) to Adobe After Effects through MCP, enabling scripted control of compositions, layers, effects, keyframes/graph easing, presets, markers, audio levels, waveform analysis, and effect discovery via a live bridge panel.
Project management your AI can actually run — connect Claude, ChatGPT, Cursor & Codex to one board over MCP.
AIOProductOS spine over MCP — customers, revenue, feedback, work, analytics on one typed record.
The memory layer for AI coding tools. Local-first, semantic, 9 MCP tools with consolidation and project scoping. Works with Claude Code, Cursor, Windsurf & any MCP client.
One MCP server for Claude, ChatGPT & Gemini — wraps your ERPs, CRMs, APIs and knowledge base into a single governed endpoint.
Run your field service business from Claude: answer calls, book and dispatch jobs, build estimates, chase invoices. Built for HVAC, plumbing, electrical, roofing and pest control.
Complete Swiss accounting integration for Bexio via MCP. Works with Claude Desktop, n8n, and any MCP client. 221 tools for invoices, contacts, projects & more.
Persistent memory MCP server for Claude Desktop — remembers context, time, and topics across sessions
An MCP extension for the Claude Desktop application that enables automation and integration.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
